Hello, dear friend, you can consult us at any time if you have any questions, add WeChat: daixieit

MBA 2023-2024 Analytics

Individual Assignment

Core Course 1 Credit

This assignment consists of three parts.  Parts  1, 2 and 3 are worth 40,  10 and 50  marks, respectively.

Your report should be submitted as a PDF document, consisting of no more than 1,500 words. The word count covers the main body of text, including in-text citations and direct quotations, and excludes equations, figures, charts, tables, footnotes, reference list and bibliography.

Please do not include an appendix, and please do not submit an Excel spreadsheet.

Part 1 – Umbrella

Umbrella is an agrochemical company that produces and distributes worldwide. Currently, they are restructuring the production of one of their fertilizers, named Growell. To produce Growell efficiently, Umbrella uses the batch approach, in which a certain number of litres is produced at one time. This reduces setup costs and allows Umbrella to produce Growell at a competitive price. Unfortunately, Growell loses some of its properties with time.

Umbrella produces Growell in batches of 20,000 litres, 30,000 litres and 40,000 litres. For simplicity, assume that in the coming three-month planning period, Umbrella will sell either 20,000 litres, 30,000 litres or 40,000 litres. More specifically, using historical data, Umbrella estimates that there is an equal probability of selling 20,000 litres, 30,000 litres or 40,000 litres.

The question Umbrella is facing is how many litres to produce of Growell in the next batch run.  Growell  sells  for  £10  per  litre.  Manufacturing  costs  amount  to  £5.50  per  litre,  and handling costs and warehousing costs are estimated to be £0.50 per litre. Umbrella allocates advertising costs to Growell at £1.50 for each litre that it produces. If Growell is not sold within three months, the fertilizer loses some of its properties. It can, however, be sold at a salvage value, estimated to be £5 per litre. Furthermore, Umbrella has guaranteed to its suppliers that there will always be an adequate supply of Growell. If Umbrella does run out, it has agreed to purchase a comparable fertilizer from a competitor at £12 per litre. No handling or warehousing costs are incurred on fertilizer provided by the competitor.

(a) Use a decision tree to identify the best decision for Umbrella. (10 marks)

(b) How does your answer to part (a) change with respect to changes in the salvage value and demand probabilities? (20 marks)

(c) What is the most that Umbrella should pay for information regarding the level of demand? (10 marks)

Part 2 – Virus Testing

(a) Obtain the reported sensitivity and specificity for a test for COVID-19. Interpret these values. (Choose a test for which neither the sensitivity nor specificity is 100%.) (5 marks)

(b) Calculate the positive and negative predictive values for the test you identified. State any assumptions that you need to make. (5 marks)

Part 3 – Shine Bright Like a Diamond

A  regression  model  is  required  to  support  the   pricing  of  individual  diamonds.  The  file ShineBright.xlsx contains the  individual selling  price and various characteristics of  10,000 diamonds. Here is a description of the variables in the dataset:

Price

Price in US dollars

Carat

Weight in carats

Cut

1 (worst) to 5 (best), representing AGS grades fair, good, very good, premium and ideal

Colour

1 (worst) to 7 (best), representing GIA grades J to D

Clarity

1 (worst) to 8 (best), representing GIA grades I1, SI2, SI1, VS2, VS1, VVS2, VVS1 and IF

Table

Width of the top of the diamond as a percentage of its widest diameter

Length

Length in mm

Width

Width in mm

Depth

Depth in mm

DepthPercentage

Depth divided by average of Length and Width

(a) Split the data for the 10,000 diamonds into a training dataset and a testing dataset. Discuss the motivation for splitting the data in this way. (5 marks)

(b) Using only the training dataset, estimate a regression model. Make sure to describe the steps of your modelling approach, and to interpret your model. (35 marks)

(c) Evaluate the accuracy of your model from part (b) in terms of its ability to predict the price of the diamonds in the testing dataset. Compare this with the accuracy of at least one other regression model estimated using the training dataset. (10 marks)